[binalpatel]

The fact that the original, unmodified article referred to data engineers as "janitors" pretty much says it all.

It's very analogous to front-office and back-office work in Investment Banking. "Data Scientist" are the front-office, with all the prestige, and "Data Engineers" are the back-office, doing a lot of the heavy lifting without nearly as much recognition.

In my opinion there shouldn't be a delineation. You shouldn't be a data scientist if you can't gather, process, and clean up your own data.

[biztos]

Ideally you'd have a symbiosis, and each side would recognize the importance of the other.

Even if you require your data scientists to be able to do engineering work, it's probably way more efficient to have some good generalist Software Engineers doing all the "pre-math" work and freeing your statisticians up for what they're (hopefully) good at.

Plus as a side effect, your software will probably be better.